Rack and Pinion drive

Has anyone else tried out the herringbone rack and pinion system on their smart rap?

I tried out the GT2 belt system using some of the STLs updated on github. The X axis belt kept slipping off of the tensioner bearing, so I replaced the X axis belt with the printed herringbone rack and pinion. My Y axis is still setup as a GT2 belt. I must say that the rack & pinion is extremely smooth! I cannot tell any difference in the prints between the x- and y-axes! I am going to switch my Y axis to rack and pinion as well.

I updated the pinion with a slot for an M3 nut and M3 set screw to hold it against the shaft. This pinion fits a little better than the keyed version.

Thingiverse link to CellJeffe's smartrap parts

I have not tried it yet, but I have given real good thought about printing some once I get the whole bed accurately calibrated. I have a set of belts coming, but I'm planning to make a second smartrap to play around with.

I've tried to print the herringbone Rack several times though Cura and slicer somehow act up and i get scewed prints. So i gave up on it and will try again later....
I tried to print diagonal on the bed to be able to print the largest one.

It's great to hear that the R&P runs so smooth!

You don't have to print them full size. I was thinking of making 2 pieces at 1/2 the size I need and use all-tread through the middle hole to join them to make one piece

I am using the split rack from the SmartRap github models - with a 5mm threaded rod through the middle and 5mm nuts on either end of the rack. I purposefully printed some test cubes and thin walled boxes so that the rack joint was directly in the middle of the printed object. I could not see any sign of the joint between the two rack halves of any of the printed objects!

I also switched to 1 perimeter and outer perimeters first in slic3r. My prints are now very accurate.

Two warning about the rack STL file on the smartrap github:
1. Print with the gear sideways. Most printers can handle the slanted overhang angles of the double herringbone.
2. One of the flat sides of the STL file is not actually flat! And the orientation of the rack80.stl on github has the curved side down which makes a bad print. Take a look layer-by-layer after slicing the racks. If you see a gap on the bottom, flip it over and re-slice.

I have been meaning to make an inventor version of the rack like I did for the pinion to correct this problem. Maybe over the weekend.

I find it amazing how much Smartriendz have been able to do with Blender -- horrible, clunky, inaccurate Blender!

Edited 1 time(s). Last edit at 07/02/2014 11:04AM by CellJeffe.

Yes, the rack&pinion works really good ! I can tell from my productions smartraps, that are normal stock smartrap and have printed now 110 entire kits ! Ihave now 2 with GT2, for practical reasons, but i can't tell which one would work better. I put some vaseline on the rack and it works reaaaally smooth ! I'm happy to have all three systems integrated now ( fishing line, r&p and GT2).

About the model not flat, it was for the X axis ( the one on top ) with 8mm rods.. It's really tight up there ! With 6mm rods and 049 version, the out of opens cad version is great .

Thank you for blender I've learned it and used it for 3d animation before, so i knew it well. I know it's quiet..surprising, for someone coming from other softwares Whe you know it, it works not that bad . I use booleans a lot to have something a little like CAD program. But i see the limits now.
Learning free cad actually I think Smartrap 2 will be in free cad or openscad fully .

---

the Smartrap project

[smartfriendz.com]
[www.thingiverse.com]
[reprap.org]
[github.com]
doc assemblage: [reprap.org]
NEW: Forum smartfriendz: [smartraptalk.smartfriendz.com]

I am running r&p aswell, and so far I can't complain, just run perfectly!
Only note i have to di is the fitting on the x axis, the rack is way too near to the bearings holder, would be nice to redesign end xis holders and pinion in order to have some more space between the rack and the bearing holder!

Yes, it's really too tight there. It's in the work to make more space. I was waiting because i wanted to do a fixed axis with head moving on it.. but it's not very good for r&p solution.

I guess i will do all.. more space for actual design, and an optional design with fixed axis for fishing line and GT2. The benefit of fix axis is that we can make it more solid, and we could put the smartrap against the wall , less space on the desktop ..

Best for me is to have the options.. anyone can choose the best for him. It's the goal of the smartrap after all

---

the Smartrap project

[smartfriendz.com]
[www.thingiverse.com]
[reprap.org]
[github.com]
doc assemblage: [reprap.org]
NEW: Forum smartfriendz: [smartraptalk.smartfriendz.com]

I've been using the shorter 0.48/0.49 bearing holders for the X axis. There is plenty of clearance now between the bearings and rack/pinion. In fact the rack fits much better there than the GT2 tensioning bearing!

ho yes. you're right. in this version i've made some place .. as we don't use an extra part to hold the bearings It's true i've used the common racks out of opens cad with this.. i forgot so much changes lately ..

---

the Smartrap project

[smartfriendz.com]
[www.thingiverse.com]
[reprap.org]
[github.com]
doc assemblage: [reprap.org]
NEW: Forum smartfriendz: [smartraptalk.smartfriendz.com]

i also use rack and pinion . i don't know why people want to go with gt2 belt !
rp is so smooth!
manage to make one 235 mm long today (tuned version )
[dl.dropboxusercontent.com]

Wow very nice print. That's how you use the whole bed travel

Hi, all

I built a "finish line" version works fine but slip occasionally, so i decided move to "R&P" version !! but i got trouble in step/mm & auto level(with servo), Did i need modify "configration.h" ??

thanks all

smartrap is great!!

I don't know the right values for R&P, but In configration.h adjust the values of

#define DEFAULT_AXIS_STEPS_PER_UNIT

Someone in another forum mentioned "sinusoidal" herringbone gears. That has me intrigued. Presumably, the helical gear mesh path develops sinusoidal error as the meshing teeth progress along the slant. A sinusoidal mesh pattern might compensate -- double it to a herringbone pattern to cancel the axial thrust. Unfortunately, the only thing I can find on sinusoidal herringbone gears are a set of patents from 1920. Unfortunately, he does not go into any advantages it might have.

Process for cutting gear wheels
Sinusoidal gearing

Quote
CellJeffe
Someone in another forum mentioned "sinusoidal" herringbone gears. That has me intrigued. Presumably, the helical gear mesh path develops sinusoidal error as the meshing teeth progress along the slant. A sinusoidal mesh pattern might compensate -- double it to a herringbone pattern to cancel the axial thrust. Unfortunately, the only thing I can find on sinusoidal herringbone gears are a set of patents from 1920. Unfortunately, he does not go into any advantages it might have.

Process for cutting gear wheels
Sinusoidal gearing

Why not use a straight rack? That would overcome that problem. Easier to print out too.

---

[regpye.com.au]
"Experience is the mother of all knowledge." --Leonardo da Vinci

Quote
regpye
Why not use a straight rack? That would overcome that problem. Easier to print out too.

A straight involute gear (straight rack) generates positional error as the contact is transferred stepwise from tooth to tooth -- so it produces bad resolution. A helical gear maintains more continuous contact that slides along the slope as motion is transferred from tooth to tooth. They get much better resolution, but a helical gear adds an axial thrust to the gear. A double helical (herringbone) gear has the nice smooth continuous motion and resolution of the helical gear, but the second slope cancels the axial thrust of the first. Herringbone cuts are just much harder to machine - so are not used that often in the "real world". But printing is another story. My smartrap can easily handle the slight overhang on each layer as it prints the helical slope of the gear.

I was just wondering if there was something better than a herringbone gear with two linear angles. I am not a gear expert.